Method and system for imaging multi-pillar samples

Abstract

To provide a method capable of adjusting a plurality of pillar samples, performing tomographic imaging, and reconstructing a 3D image in a short time.SOLUTION: A method includes: providing a multi-pillar sample including at least a first pillar and a second pillar parallel with the first pillar; directing a charged particle beam to the first pillar; imaging the first pillar at a plurality of rotational positions by rotating the multi-pillar sample about a first pillar axis of the first pillar; directing the charged particle beam to the second pillar; and imaging the second pillar at a plurality of rotational positions by rotating the multi-pillar sample about a second pillar axis of the second pillar.SELECTED DRAWING: Figure 1A

Description

[Technical Field]

[0001] This field concerns tomography. [Background technology]

[0002] An example of thin-section tomography involves using a transmission electron microscope (TEM) to acquire a series of images of a sample at various tilt and rotation positions. For example, as shown in FIGS. 1A-1C, a thin-section sample 100 can extend within a sample plane 102 to form a thin, planar layer presenting multiple thin regions for tomographic imaging. The sample 100 can be positioned within a sample carrier 104 held by a translation stage 106 within the field of view of an electron beam 108. The electron beam passes through the sample 100 and is received by an image sensor 110 to form an image. To generate a series of images for tomographic reconstruction, the translation stage 106 tilts the sample 100 through a series of rotation angles. For example, the sample 100 can be imaged at 0 degrees, where the sample plane 102 is perpendicular to the axis 112 of the incident electron beam, as shown in FIG. 1B, and through a series of angular increments provided by the translation stage within a predetermined ± rotation range of the sample plane 102 from the 0 degree position. Images are captured at various tilt positions, stored, and used to reconstruct a tomogram or 3D image of the sample.

[0003] For thin samples, this tilt range is often limited, for example, to -70 / +70 degrees, as shown in Figure 1C, by the sample carrier 104 supporting the sample 100. Such limitations are usually associated with the edges of the carrier or other support structure of the sample interfering with the electron beam used to image the sample. Another problem that arises with thin samples is that as a thin sample is tilted to higher tilt angles, the effective thickness of the sample increases, as indicated by the arrows in Figures 1B-1C. A larger effective thickness attenuates the transmitted electron beam, reducing image quality and, therefore, the quality of tomographic reconstructions.

[0004] Thus, many problems associated with thin section tomography remain. The tilt range for rotating a thin section sample is typically limited, for example, by edge obstructions and associated imaging effects of the carrier used to hold the thin section sample, or by the effective thickness of the thin section sample changing as the rotational position of the sample changes through the tilt range of the tomographic acquisition, as discussed above. This limited tilt range leads to limited information obtained from the sample and reduced quality tomographic reconstructions. Therefore, a need exists for improved tomographic techniques. Summary of the Invention

[0005] According to one aspect of the disclosed technology, a method includes providing a multi-pillar sample including at least a first pillar and a second pillar parallel to the first pillar; directing a charged particle beam toward the first pillar; imaging the first pillar at multiple rotational positions by rotating the multi-pillar sample about a first pillar axis of the first pillar; directing the charged particle beam toward the second pillar; and imaging the second pillar at multiple rotational positions by rotating the multi-pillar sample about a second pillar axis of the second pillar. Some methods include reconstructing a 3D image of the first pillar and the second pillar. In some reconstructions, the 3D image covers the entire angular span of the first pillar and the second pillar. In some examples, the range of the multiple positions is at least 170 degrees. In some examples, the range of the multiple positions is at least 80 degrees. In some examples, the range of the multiple positions is at least 180 degrees. Some examples include translating the multi-pillar sample along the first pillar axis to rotationally image another section of the first pillar before moving the multi-pillar sample to image the second pillar. In some examples, imaging the first pillar includes rotating the multi-pillar sample to a selected position of either a +90 degree position or a -90 degree position, and generating an image at the selected position where obstruction by the second pillar or other pillars of the multi-pillar sample is reduced based on an angled linear arrangement of the multiple pillars of the multi-pillar sample relative to the 0 degree position. In some examples, the multi-pillar sample includes a substrate defined by a length, a width, and a height, and the first pillar and the second pillar extend from the substrate along the height of the substrate, are spaced apart, and are sequentially positioned at different positions on the length of the substrate. Some multi-pillar sample substrates are sample carriers. Some examples include attaching multiple pillars to a substrate to form a multi-pillar sample. In some examples, the distance between the first pillar and the second pillar is greater than 10 times the diameter of either the first pillar or the second pillar.Some examples include directing a focused ion beam to mill a raw sample substrate to form a multi-pillar sample.

[0006] According to another aspect of the disclosed technology, an apparatus includes an imaging system configured to direct a charged particle beam at a multi-pillar sample including at least a first pillar and a second pillar; a translation stage configured to translate and rotate the multi-pillar sample about a plurality of different pillar axes of the multi-pillar sample; and a processor and memory coupled to the sample stage and the imaging system, the memory including code that, when executed by the processor, causes the sample stage to rotate the multi-pillar sample about a first pillar axis of the first pillar and to rotate the multi-pillar sample about a second pillar axis of the second pillar, and causes the imaging system to direct an imaging beam through the first pillar at a plurality of rotational positions about the first pillar axis and to direct the imaging beam through the second pillar at a plurality of rotational positions about the second pillar axis. Some examples include an imaging sensor for detecting charged particles transmitted through the first pillar and the second pillar, and the memory includes further code that, when executed by the processor, causes the imaging sensor to detect a plurality of first images of the first pillar at a plurality of rotational positions about the first pillar axis and a plurality of second images of the second pillar at a plurality of rotational positions about the second pillar axis. In some examples, the memory includes further code that, when executed by the processor, causes the processor to reconstruct 3D images of the first pillar and the second pillar based on the plurality of first images and the plurality of second images, respectively. In some examples, the multi-pillar sample comprises a substrate defined by a length, a width, and a height, the substrate extending in a plane defined by the length and width, the first pillar and the second pillar extending parallel to each other from the substrate along a height of the substrate, being spaced apart and consecutively positioned at different positions along the length of the substrate. Some examples further include a sample carrier, the substrate being the sample carrier. In some examples, the first pillar axis and the second pillar axis form a pillar plane that is not parallel to a plane defined by the length and width of the substrate.In some examples, the memory includes further code that, when executed by the processor, causes the sample stage to move the multi-pillar sample such that the charged particle beam is directed at the first pillar or the second pillar. In a representative example, the first pillar and the second pillar have a tip thickness of less than 400 nm, a base thickness of more than 400 nm, and a length of more than 1 μm. In a further example, the first pillar and the second pillar have a tip thickness of less than 600 nm, a base thickness of more than 400 nm, and a length of more than 1 μm.

[0007] The foregoing and other objects, features, and advantages of the disclosed technology will become more apparent from the following detailed description, which proceeds with reference to the accompanying drawings. [Brief explanation of the drawings]

[0008] [Figure 1A] 1 shows an image of a thin lamina sample rotated by a translation stage considering a transmission electron beam. [Figure 1B] 1B are side views of the sample of FIG. 1A at 0 and 70 degrees, respectively. [Figure 1C] 1B are side views of the sample of FIG. 1A at 0 and 70 degrees, respectively. [Figure 2A] FIG. 1 is a perspective view of a tomographic sample with pillars extending therefrom. [Figure 2B] 2B are side views of the sample of FIG. 2A at 0 and 70 degrees, respectively. [Figure 2C] 2B are side views of the sample of FIG. 2A at 0 and 70 degrees, respectively. [Figure 3] FIG. 1 is a perspective view of a tomographic pillar sample held by a sample carrier and translation stage in a tomographic environment. [Figure 4] 1 shows various perspective side views of the geometries of the fault pillars extending from their respective bases on the left to their tips on the right. [Figure 5] FIG. 2 is a perspective side view of a tapered pillar. [Figure 6]FIG. 1 is a perspective view of a multi-pillar sample placed in a tomographic environment. [Figure 7] FIG. 1 is a perspective view of a multi-pillar sample held by a sample carrier and translation stage in a tomographic environment. [Figure 8] FIG. 1 is a side view of two pillars of a multi-pillar sample. [Figure 9] FIG. 1 is a side view of five pillars of a multi-pillar sample. [Figure 10] Side view of a multi-pillar sample with five pillars arranged along a line at a slight angle to the normal of the incident electron beam. [Figure 11] 1 is an image of a multi-pillar sample obtained by electron microscopy. [Figure 12] 1 is an image of a multi-pillar sample obtained by electron microscopy. [Figure 13] FIG. 1 is a schematic diagram of an electron microscope with an optional focused ion beam column. [Figure 14] FIG. 10 is a side view of an irregularly shaped multi-pillar sample and sample carrier. [Figure 15] FIG. 1 is a side view of a multi-pillar sample in which pillars are arranged in a grid pattern. [Figure 16] 1 is a flowchart of a method for fabricating a multi-pillar sample. [Figure 17] 1 is a flowchart of a method for tomographic imaging of a multi-pillar sample. [Figure 18] FIG. 10 is a perspective view of an additional multi-pillar sample. [Figure 19] FIG. 10 is a perspective view of an additional multi-pillar sample. [Figure 20] 10 is a flowchart of an additional method for fabricating a multi-pillar sample. [Figure 21] 10 is a flowchart of an additional method for fabricating a multi-pillar sample. DETAILED DESCRIPTION OF THE INVENTION

[0009] One approach to addressing these issues may involve using a sample pillar 200 extending from a sample substrate 202, as shown in Figures 2A-2C, instead of just a thin 2D thickness of sample material. The pillar 200 extends along a pillar axis 204 perpendicular to the axis 206 of an incident imaging beam 208, such as an electron beam or other particle beam. A series of tilt images can then be acquired as the sample pillar 200 is rotated around the pillar axis 204 to various tilt angles. This can reduce or eliminate edge disturbances caused by the sample carrier, thereby allowing a full 360-degree range, or a full +90 / -90-degree range, to be used to acquire tomographic images of the sample. For example, Figure 2B shows the sample plane 210 of the sample substrate 202 positioned at a 0-degree position perpendicular to the incident electron beam 208, while Figure 2C shows the sample plane 210 positioned at a 70-degree position relative to the 0-degree position. A carrier can be used to provide the pillars with spatial clearance for imaging, allowing the pillars to replace thin flake grids that may cause edge obstructions. For example, the sample pillars can extend into the field of view 212 of the imaging beam 208 without the edges of the carrier substantially interfering with the image formed or collected.

[0010] Another example of avoiding edge interference is shown in Figure 3. Sample 300 includes a bulk substrate 302 and a pillar 304 extending therefrom along a pillar axis 306. Sample 300 is positioned on the base of a carrier 308, which is held by a translation stage 310, also referred to as the sample stage.

[0011] The sample 300 can be secured to the carrier 308 in various ways. In a typical example, the sample 300 is attached to the carrier 308 by freezing it from a liquid state in a vitrification process, such as using high-pressure freezing (HPF). The freezing process can fix the vitrified sample material to the surface of the carrier 308, which can also define a sample plane for the sample, as the sample material generally conforms to the surface of the carrier 308. Adhesion between the frozen material and the carrier can be maintained by appropriate selection of the carrier material, carrier geometry, and / or additives to the sample prior to vitrification. For temperature-controlled or room-temperature samples, bulk sample material can also be attached to the carrier 308 in other ways, such as chemical welding, microwelding, adhesives, or redeposition. A gas injection system (GIS) can be used, for example, in instances where a precursor to the sample is destroyed and a beam scans a portion of the sample to create a weld. The lift-out process can be used to transfer sample material before or after pillars are formed in substrate 302, for example, by using a micromanipulator arm to pick up and place sample 300 on carrier 308 within the FIB / SEM microscope chamber. In a further example, pillars 304 can be attached directly to carrier 308 without a bulk substrate. In some examples, sample 300 and carrier 308 can be the same.

[0012] The translation stage may be configured to rotate the carrier 308 about an axis aligned with the pillar axis 306 so that the incident imaging beam 312 is directed through the pillar 304 at various tilt positions about the pillar axis 306 to generate tomographic image slices of the pillar 304.

[0013] The pillars can be configured in a variety of shapes. For example, each pillar can have a height / length defined along the pillar axis and at least a width / thickness defined in a plane perpendicular to the pillar axis. The pillars have an aspect ratio greater than 1. While Figures 4A-4D show different pillar shapes, such as cylindrical 400A, conical 400B, tapered 400C, extended 400D, rectangular 400E, and irregular 400F, it will be understood that various other shapes can be provided for tomographic pillar imaging. The shape of the pillar can be defined in relation to the milling process used to produce the pillar, such as focused ion beam (FIB) milling. Thus, rather than imaging a large and thin sample, a pillar-shaped structure can be imaged.

[0014] Referring to FIG. 5 , pillar 500 extends along pillar axis 501 from base region 502 bonded to bulk substrate 504 and is tapered along all or a portion of its length before terminating at pillar tip region 506. Thus, the thickness or diameter of the pillar may vary along the length of different pillar examples, and various portions of the pillar can be selected for imaging. In some examples, the pillar portion selected for imaging is constrained based on the pillar thickness and imaging capabilities of the incident probe beam, such as particle selection, electron energy, wavelength, etc., with increasing thickness resulting in decreased transmission of the imaging beam through the sample due to absorption, scattering, etc. In some examples, the tapered pillar portion used for transmission imaging can have a thickness of tens to hundreds of nanometers, such as 10 nm, 20 nm, 50 nm, 100 nm, 200 nm, 300 nm, 400 nm, 500 nm, or 600 nm, although other thicknesses are possible. The base of the pillar can have various thicknesses to support the pillar extension, such as 400 nm, 500 nm, 600 nm, or 1 μm. Exemplary pillars can have various lengths along the pillar axis, such as 100 nm, 200 nm, 500 nm, 1 μm, 2 μm, 5 μm, or 10 μm, which can be determined by the material or sample type (e.g., cryogenic, biological, semiconductor, etc.) or FIB milling process. Generally, larger aspect ratios, such as longer pillars and sufficiently narrow thicknesses, are preferred to provide additional sample material for tomographic imaging and to provide greater distance from potentially interfering objects, such as attached sample substrates, carriers, or translation stages. In some instances, samples can be formed by cutting the pillars and transferring them from one substrate to another, e.g., by attaching the pillar base region to the substrate. In preferred instances, pillars are formed by milling a bulk substrate, which can streamline sample preparation. In some examples, the bulk substrate may have multiple separate substrates joined together, each with one or more pillars milled into each separate substrate.

[0015] While TEM acquisition characteristics may be improved, the use of pillars in tomographic imaging as an alternative to thin-section imaging is less desirable because the amount of material available for imaging and corresponding analysis may be significantly less than a thin, lamina sheet extending substantially in the two-dimensional sample plane. Multiple pillars may be imaged to analyze more material or multiple locations in a bulk substrate. However, preparing and imaging each pillar individually can be time-consuming, especially if sample preparation and sample imaging are performed in different tools. For example, pillars may be prepared in a FIB-SEM system, and the sample may be imaged in a TEM. In some instances, the sample may be vitrified in yet another tool.

[0016] The above problems can be solved by preparing and imaging a multi-pillar sample. In one example, a multi-pillar sample includes multiple pillars extending from a substrate. The pillars may be prepared by milling a bulk substrate using a FIB. Alternatively, the pillars may be individually prepared using a FIB and then attached to the substrate. In another example, a multi-pillar sample includes multiple pillars attached to a sample carrier. Each pillar may be prepared using a FIB and then attached to the sample carrier.

[0017] Tomographic images of one or more cross sections of each pillar of a multi-pillar sample can be acquired by rotating the sample relative to its pillar axis via a translation stage. The sample's rotational range may be at least 80, 90, 100, 110, 120, 130, 140, 150, 160, or 170 degrees. In some examples, the sample's rotational range is at least 180 degrees. However, while imaging a selected pillar, the imaging beam may be blocked by other pillars at certain rotational positions. The effect of beam blockage on tomographic imaging may be reduced or minimized by specific pillar placement. Multiple pillars may be positioned at different locations along the length of the substrate / sample carrier. Multiple pillars may be parallel to one another and have the same height. In various multi-pillar examples disclosed herein, the degree to which pillars are considered parallel may vary depending on manufacturing tolerances or pillar geometry, e.g., the axes of the respective pillars may form an angle of 0.5 degrees or less, 1 degree or less, 2 degrees or less, 5 degrees or less, 10 degrees or less, 20 degrees or less, etc. In one example, the pillars are positioned approximately linearly along a line. The line may run along the length of the substrate / sample carrier. Alternatively, the line may have a small angle (e.g., less than 10 degrees) from the length of the substrate / sample carrier.

[0018] Referring to FIG. 6, an example of a multi-pillar sample 600 is disclosed. The multi-pillar sample 600 includes multiple pillars 602a-602e extending from a bulk substrate 606 along respective axes 604a-604e. The axes 604a-604e may be parallel to one another. The substrate is defined by a length (Y-axis), a width (Z-axis), and a height (X-axis). The length and height of the substrate define a sample plane 616. The pillars 602a-602e are spaced apart and arranged consecutively along the length of the substrate 606 by a distance separating adjacent ones of the pillars 602a-602e. The sample 600 may be positioned on a sample carrier 608 coupled to a motion stage 609 in a tomography environment 610 of a transmission electron microscope, for example, under vacuum and / or cryogenic conditions. The sample 600 and pillars 602a-602e and tomographic environment 610 are shown positioned relative to a coordinate system 611. Coordinate system 611 can be understood as two separate coordinate systems: one for the sample 600 and pillars 602a-602e and one for the tomographic environment 610. As shown, the coordinate systems are aligned with the sample 600 in an aligned position in the tomographic environment 610, e.g., with the sample plane 616 perpendicular to the imaging beam axis 612 of the imaging beam 614.

[0019] In some examples, the pillars 602a-602e may also be spaced apart and consecutively arranged such that the pillars 602a-602e are distributed along the imaging beam axis 612 of the imaging beam 614, e.g., along the z-axis in FIG. 6. Such distributed spacing may be understood with reference to a sample plane 616 that is perpendicular to the imaging beam axis 612. The sample plane 616 may be parallel to the XY plane and one or more reference planes of the bulk substrate 606, such as the top and / or bottom surfaces 615a, 615b. For example, the sample 600 may be positioned on the sample carrier 608 such that the sample plane 616 is parallel to the lateral extent or reference plane of the sample carrier 608 and / or perpendicular to the imaging beam axis 612, with the translation stage 609 holding the sample carrier 608 in the 0 degree position. As shown, imaging beam 614 is directed through pillar 604c to imaging sensor 617, such as a pixelated array of CCD or CMOS sensors.

[0020] In a representative example, a series of spaced pillars 602a-602e are aligned in a line 618. As shown in FIG. 6, axes 604a-604e are spaced apart and extend parallel to one another and to the sample plane 616. Axes 604a-604e may intersect line 618 to form a reference pillar plane 620 that is rotated about axis 604c and forms a small angle with respect to the sample plane 616. If other size and spacing criteria are met (discussed further below), the spacing distribution along the z-axis associated with a small angle allows the sample 600 to be rotated to positional extremes of + / -90 degrees while avoiding shadowing or obstruction by adjacent pillars during image acquisition of a particular one of the pillars 602a-602e. Exemplary small angles can be approximately 0-2 degrees, 0.5-1.5 degrees, 1-3 degrees, 2-4 degrees, 4-6 degrees, 6-8 degrees, 8-10 degrees, or more. In some examples, the interfering angle ranges can be avoided during acquisition or discarded during post-acquisition tomographic image reconstruction.

[0021] In some examples, a small angle between the planes 616, 620 may be zero, and certain images captured at extreme positions may be discarded if obstructions are present. In such cases, the line 618 may be parallel to the sample plane 616 and / or perpendicular to the imaging beam axis 612 with the sample 600 held at the 0 degree position.

[0022] In some examples, line 618 may be parallel or at an angle to the sample plane 616 or another reference plane of the sample 600 or carrier 608, and the motion stage 609 can adjust the rotation and translation of the sample 600 so that line 618 is positioned at a small angle to a line perpendicular to the imaging beam axis 612, providing a zero degree position for imaging. Similarly, the motion stage 609 can adjust the rotation and translation of the sample 600 so that line 618 is perpendicular to the imaging beam axis 612, i.e., not at a small angle.

[0023] During image acquisition, sample 600 is rotated by translation stage 609 around the pillar axis of the pillar being imaged, which may correspond to the eucentric axis of translation stage 609. For example, as shown in FIG. 6, sample 600 can be rotated around pillar axis 604c to acquire a series of tilt images of pillar 602c. Due to the spaced, sequential arrangement, when the sample is rotated near the +90-degree or -90-degree position, adjacent pillars 602b, 602d cause little interference with image acquisition of pillar 602c. After the image of pillar 602c is acquired, translation stage 609 can reposition sample 600 to rotate around a new pillar axis, such as pillar axis 604d, to acquire a series of tilt images for pillar 602d. Again, the spaced, contiguous arrangement allows adjacent pillars 602c, 602e to provide little interference with image acquisition for pillar 602d over the entire +90 / -90 degree tilt range.

[0024] The spacing between adjacent pillars is preferably substantially greater than the thickness of the pillars 602a-602e in the imaging region of interest, e.g., near the pillar tips or in sufficiently tapered sections to allow imaging beam transmission. Increasing the distance between adjacent pillars ensures that any overlap occurring in the path of the imaging beam 614 is limited to a small angular range. For example, the distance between adjacent pillars (which need not be identical between each pair of adjacent pillars) can range from 20-50 μm, 50-100 μm, 100-200 μm, 200-400 μm, or 400+ μm, although other distances may be suitable. In the specific example shown in Figure 8, two adjacent pillars with a diameter of 400 μm in the imaging cross section are 50 μm from each other, resulting in an angular overlap of 0.46 degrees. In many tomographic acquisitions, such overlap can represent negligible information loss. For example, in many tomography applications, tilt step increments can be selected in the range of approximately 1-3 degrees, resulting in zero or minimal degradation of image quality for images acquired angularly close to obstructions or overlaps. Exemplary angular overlaps can include less than approximately 0.1, 0.25, 0.5, 1, 1.5, 3, or 5 degrees. In some TEM system applications, the carriers can have an available width of less than approximately 750 μm, allowing for a 50 μm spacing between pillars and a continuous distribution that allows for up to 15 pillars per carrier. In various examples, the ratio of adjacent pillar distance to pillar thickness can be at least approximately 20 or more, 30 or more, 50 or more, 100 or more, 150 or more, 200 or more, or 500 or more. Exemplary ratios can be determined by the sample type and preparation, including available milling processes and the sample's fracture strength.

[0025] The linear arrangement of the spaced series of pillars 602a-602e also reduces the amount of overlap during image acquisition of the tilt series because the closer the pillars 602a-602e are aligned along line 618, the smaller the size of any angular shadows cast by adjacent pillars. Therefore, by arranging the spaced series linearly where any overlap occurs, it occurs on all pillars at once, and no more information is lost than if there were only two pillars, with the information loss defined solely by the adjacent pillars. Furthermore, because images are typically acquired relative to a specific pillar, adjacent pillars that are angularly close enough to cause interference are also defocused, further reducing the impact of adjacent pillars on image quality. It can generally be difficult to arrange the pillars 602a-602e so that line 618 is perfectly straight. Therefore, the described line and linear arrangement may deviate from the ideal line by various extents or amounts. For example, an exemplary line including three or more pillars may have a thickness or variation corresponding to 0.5, 1, 5, 10, or 50 pillar diameters. In some examples, the distance of each pillar from the line is within 10 or 50 pillar diameters. In some examples, the third pillar of a group of three pillars is sufficiently distant from the line that intersects the other two pillars such that a separate line can be defined between each of the two pillars and the third pillar.

[0026] As discussed above, in some examples, the pillars 602a-602e may be spaced apart along the z-axis. Such spacing may require a small translation of the sample 600 by the translation stage 609 in the z-direction along the imaging beam axis 612, where the sample 600 is repositioned to acquire a series of tilt images for a newly selected pillar that replaces the previous pillar as the rotation axis. By providing spacing along the direction of the imaging beam axis 612, along with lateral spacing, i.e., perpendicular to the imaging beam axis 612 at the 0-degree rotation position, the full -90 / +90 field of view range can be reached for each of the pillars 602a-602e. While some information may be lost between -90 and +90, it may be beneficial to acquire image information at the angular extremes. It may also be beneficial to acquire images from as many different image projections as possible. For example, if a particular projection is more distinct from other projections, acquiring that particular projection may provide more comprehensive information for the 3D tomographic reconstruction compared to fewer or more projections in a smaller or limited angular range. In some examples, other rotation ranges may be used based on, for example, the sufficiency of information for a particular sample, motion stage constraints, housing constraints, sample carrier geometry, etc. Rotation ranges of ±180 degrees, ±150 degrees, ±120 degrees, ±100 degrees, ±95 degrees, ±85 degrees, ±80 degrees, ±70 degrees, ±60 degrees, ±50 degrees, or ±40 degrees may also be used in some examples. Some rotation ranges may include asymmetric ranges, e.g., +90 / −40, +95 / −85, +40 / −90, +70 / −75, etc.

[0027] As discussed above, multi-pillar samples can be generated using various FIB techniques, and different sample types can define the geometry, spacing, and other characteristics of the pillars generated. In some examples, pillars are generated from bulk samples prepared by high-pressure freezing (HPF). By generating more pillars in samples where single pillars are typically used, the number of multi-pillar samples that can be created to acquire the same amount of data as a set of single-pillar samples can be significantly reduced, thereby reducing the time, cost, and difficulty of generating, staging, carrier exchange, and tomographic imaging of many separate samples. Furthermore, the required carrier exchange complicates automation of data acquisition. Furthermore, two samples are never prepared using exactly the same conditions, which can adversely affect the consistency of analytical data. Therefore, in the disclosed examples, even with standard TEM carrier-pillar sample dimensions, sample overlap interference that can occur during TEM acquisition can be negligible. At the same time, imaging throughput is significantly improved while maintaining the 360° coverage associated with the pillars.

[0028] 7 illustrates a tomography environment supporting a multi-pillar sample 700. The sample 700 includes a bulk substrate 702 and a plurality of pillars 704a-704e extending from the substrate 702. The sample 700 is disposed on a carrier 708 held by a translation stage 710. The sample 700 may be secured to the carrier 708 in a variety of manners, as previously discussed. The translation stage 710 may be configured to rotate the carrier 708 about the axis of each pillar 704a-704e such that an incident imaging beam 712 is directed through each pillar 704a-704e, for example, by sequentially acquiring images at a series of tilt angles for each pillar 704a-704e.

[0029] FIG. 8 shows a set of pillars 800a, 800b of a multi-pillar sample viewed from an end view, with the pillars extending out of the plane of the drawing along their respective axes 802a, 802b. Pillars 800a, 800b each have a diameter of approximately 400 nm and are spaced apart from one another by a distance of 50 μm. Pillar 800a is imaged through a series of tilt tomographic images by rotating it about axis 802a. As it rotates about axis 802a, the position of pillar 800 moves along arc 804. Due to the cross-sectional size of the pillars and the distance between them, each pillar overlaps the others by approximately 0.46 degrees, an exemplary amount that can avoid obstructing the tomographic images at a suitable series of tilt angle increments.

[0030] Figure 9 shows a set of pillars 900a-900e of a multi-pillar sample viewed from an end view, with the pillars extending out of the drawing along their respective pillar axes 902a-902e. Pillars 900a-900e are generally arranged along line 904 to reduce the amount of overlap or shadowing at a particular set of tilt angles caused by adjacent pillars not currently being imaged. For example, if pillar 900c rotates about axis 902c during image acquisition, lines 906a, 906b indicate the degree to which adjacent pillars 900b, 900d may interfere with imaging near a particular tilt angle, e.g., +90 degrees or -90 degrees.

[0031] FIG. 10 shows a set of pillars 1000a-1000e of a multi-pillar sample viewed from an end view, with the pillars extending out of the drawing. To expand the angular range for acquiring tomographic images of each of the pillars 1000a-1000e, the pillars 1000a-1000e are aligned on a line 1002 and spaced at different locations along the length (Y-axis) of the substrate 1005. The pillars may also be spaced along the width (Z-axis) of the substrate. When the multi-pillar sample is positioned at a 0-degree position within the imaging system as shown in FIG. 7, the XY plane of the sample may be aligned with or correspond to the sample plane, and the width of the substrate (Z-axis, vertical in FIG. 10) is aligned with the direction of the incident electron imaging beam 1004. In particular, the vertical spacing (along the Z-axis) is selected so that each pillar is shifted relative to its neighboring pillars by at least a distance corresponding to their thickness or diameter, e.g., so that the top of one pillar is not higher than the bottom of the adjacent pillar. By vertically spacing the pillars 1000a-1000e, cross-sectional images can be acquired by rotating the multi-pillar sample with the sample stage from -90° to +90° relative to the pillar axis of each pillar, including projections at the extremes of -90° and 90°. The tilt of the line 1002 relative to the sample plane can be selected based on the selected acquisition scheme. For example, the tilt can determine the angular range over which image interference occurs due to pillar overlap, and some pillar samples can benefit from reduced interference at a selected imaging angle or range of imaging angles. Therefore, an acquisition scheme and corresponding tilt can be selected to take advantage of a particular pillar material, geometry, or sample placement. For example, the tilt can be determined based on the rotation range of the sample stage, the placement of the pillars relative to the sample plane, and / or the pillar size.

[0032] Figure 11 is an image of a multi-pillar sample 1100 having three tapered pillars 1102a-c extending from a bulk substrate 1104. The pillars 1102a-c are formed by milling the bulk substrate 1104 with a focused ion beam. Figure 12 is an image of the same multi-pillar sample 1100 from an optical microscope.

[0033] FIG. 13 illustrates an exemplary imaging system including a charged particle microscope 1300 configured to generate a tomographic imaging beam 1302 of electrons and direct the imaging beam 1302 to a multi-pillar sample 1304 disposed within a sample chamber 1306. In a representative example, the housing chamber is cryogenically cooled to perform cryogenic electron tomography on the multi-pillar sample 1304. The exemplary imaging system can direct the imaging beam 1302, for example, in the form of an electron beam or other charged particle beam, through the sample 1304 to collect transmitted particles and form an image with an imaging sensor 1307. The microscope 1300 includes a translation stage 1308 configured to rotate the multi-pillar sample 1304 about various pillar axes to acquire images of a series of tilts of each of the different pillars of the sample 1304. In some examples, the microscope 1300 can include a focused ion beam FIB column 1310, which can be used to mill a bulk substrate 1312 of the multi-pillar sample 1304 to create pillars prior to tomographic imaging. The microscope 1300 may be controlled by a control environment 1314 comprised of at least one processor and memory that stores computer-executable instructions that operate the motion stage and various microscope components such that the multi-pillar sample 1304 is tomographically imaged and / or milled to produce its pillars.

[0034] In some examples, the microscope 1300 can be a FIB or FIB / SEM system in which multiple pillar samples are prepared for tomographic imaging and transferred to a TEM system. In some examples, the microscope 1300 can be a FIB / SEM or FIB / TEM in which multiple pillar samples are prepared and tomographically imaged. In further examples, an SEM system can be used to tomographically image multiple pillar samples. In examples in which multiple pillar samples are prepared and imaged in a common location, for example, in the same device or chamber, the imaging process can be simplified by avoiding realignment steps or additional motion stage manipulations.

[0035] For example, control components include desktop or laptop computers, mobile devices, tablets, logic controllers, etc. Processors include CPUs, GPUs, ASICs, PLCs, FPGAs, PLDs, CPLDs, etc., which can perform various data processing or I / O functions associated with the controller environment 128, such as control and / or data processing. Memory can be volatile or non-volatile (RAM, ROM, flash, hard drives, optical disks, etc.), fixed or removable, and coupled to the processor. The memory can provide storage capacity on one or more computer-readable media. The controller environment 1314 can also be located in a distributed manner, such that applications and tasks are performed by remote processing devices linked via a communications network. Program modules and logic can be located in both local and remote memory storage devices. In some examples, components within the controller environment 1314 need not be linked with other components.

[0036] The pillar image acquisition algorithms or steps and pillar substrate milling algorithms or steps may be embodied as software or firmware instructions executed by a digital computer. For example, any of the disclosed techniques may be implemented by a computer or other computing hardware that is part of or coupled to a TEM and / or FIB system. For example, software for implementing any of the disclosed embodiments may be stored on one or more volatile, non-transitory computer-readable media as computer-executable instructions that, when executed by one or more processors, cause the one or more processors to perform any of the disclosed techniques. Results of image acquisition, calculation, or reconstruction may be stored on one or more tangible, non-transitory computer-readable storage media (e.g., using suitable data structures or lookup tables) and / or output to a user, for example, by displaying data, images, tomographic image slices, 3D reconstructions, etc. on a display.

[0037] In some examples, tomographic reconstruction of one or more pillars can be performed. Often, the reconstruction can be performed locally and / or remotely, for example in a distributed computing environment, to improve processing speed or address cumbersome computational loads. Various tomographic reconstruction techniques can be used, including iterative approaches, search, expectation maximization, weighted backprojection, etc. Images can be stored and resampled as needed so that acquired image sequences can be tailored or adapted to standard reconstruction techniques.

[0038] FIG. 14 illustrates an end view of a multi-pillar sample 1400 including multiple pillars 1402a-1402e extending out of the plane of the drawing from a bulk substrate 1403. With the sample 1400 positioned at a 0-degree position (i.e., perpendicular to the electron beam) for tomographic imaging, the pillars 1402a-1402e are aligned with a line 1404, preferably at a slight angle relative to the y-axis. The y-axis is perpendicular to the z-axis, which is parallel to the propagation axis 1406 of the incident imaging beam 1408. The sample 1400 is coupled to a carrier 1410 supported by a translation stage 1412. One or more surfaces of the sample 1400 and / or carrier 1410 may include irregular surfaces, making alignment difficult. The translation stage 1412 may be configured to adjust the positioning of the sample 1400 such that the line 1404 and pillars are aligned for tomographic imaging. In a representative example, the line 1404 is positioned at a slight angle at the 0 degree position so that the boundary +90 and -90 degree positions can be imaged without being obstructed by adjacent pillars. In some examples, the line 1404 can be positioned at no angle at the 0 degree position or at another angle for cross-sectional imaging.

[0039] 15 illustrates a multi-pillar sample 1500 including multiple sets of pillars 1502a-1502e, 1504a-1504e, and 1506a-1506e viewed from an end view, with the pillars extending out of the plane of the bulk substrate 1508. Each of the pillar sets 1502a-1502e, 1504a-1504e, and 1506a-1506e may be positioned along a respective line 1502f, 1504f, and 1506f, such that small rotation angles during tomographic imaging may cause obstruction by adjacent pillars within the set. The pillar sets 1502a-1502e, 1504a-1504e, and 1506a-1506e may be arranged in a grid pattern with sufficient distance between pillars to reduce obstruction by pillars at various angular positions during tomographic imaging of a series of tilts. For example, the rotational position for pillar 1502a associated with an obstruction along intersection lines 1510a, 1510b, 1502f may be adjusted to avoid or reduce adverse effects on image quality, for example.

[0040] FIG. 16 is a method 1600 for producing a multi-pillar sample, which may include placing a bulk substrate on a focused ion beam (FIB) sample stage at 1602, and directing the FIB at the sample at 1604 to selectively remove material until multiple pillar extensions remain on the bulk substrate.

[0041] FIG. 17 illustrates a method 1700 for tomographically imaging a multi-pillar sample using a transmission electron microscope (TEM). Method 1700 may include, at 1702, holding a multi-pillar sample with pillars selected for imaging. The sample is preferably in a eucentric position so that rotation of the sample does not require significant adjustment to the sample after rotation. A eucentric position may generally correspond to a position (such as a point, line, or height) where tilt rotation of the sample, e.g., through rotation by a translation stage, does not cause significant displacement of the sample within the microscope's field of view that would otherwise require repositioning or realignment. At 1704, the positioning of the sample may be confirmed or otherwise aligned so that the pillars of the multi-pillar sample are positioned at a slight tilt angle from the 0-degree imaging position, thereby avoiding subsequent interference with subsequent rotations of the sample to tilt extremes, such as +90 degrees or -90 degrees. By way of example, a slight tilt angle may be achieved by preparing pillars offset from one another along the vertical direction of the substrate (as shown in FIG. 10 ) or by rotating a translation stage through a slight tilt angle. In 1706, an imaging beam is directed at a selected pillar, and a tomographic image is generated. In 1708, the sample is rotated about the axis of the selected pillar by a selected rotational increment (or continuous rotational motion), and in 1710, subsequent images of the sample at subsequent rotational positions are formed with the imaging beam. In 1716, method 1700 checks whether all rotational positions have been imaged. If the answer is “no,” the sample is rotated in 1708, and another tomographic image is acquired. If all rotational positions have been imaged, another position along the pillar axis of the selected pillar may optionally be imaged. If another position of the selected pillar is to be imaged in 1718, the sample may be translated along the rotational axis of the selected pillar being imaged, and an additional series of tomographic images of another cross-section of the selected pillar may be collected in 1712. In one example, the sample is translated along the pillar axis by operating a translation stage, and in another example, the imaging beam may be deflected along the pillar axis to image different pillar positions.After a selected pillar has been imaged, additional pillars may be imaged by translating the sample in the x, y, and / or z directions using the translation stage so that adjacent or other selected pillars of the multi-pillar sample may be rotationally imaged in 1714. In some examples, imaging rotational positions may also include imaging in various sequences, for example, by imaging the rotational position of a selected pillar, translating to a new pillar position, or selecting a new pillar and imaging the new position or pillar, and returning to different rotational increments of the initial selected pillar.

[0042] After images of a multi-pillar sample are acquired, various tomographic reconstruction techniques can be used to form a 3D reconstructed image from the acquired images, for example, covering the full angular span (360 degrees) of each imaged pillar. The pillar geometry allows for a more complete range of images to be acquired compared to thin-section samples. Due to the placement of the pillars, the acquired tomographic image for a particular pillar may have minimal or no obstruction from other pillars. Therefore, one or more missing wedges associated with obstructions from other pillars may not be present in the reconstructed 3D image. Reconstruction techniques include iterative approaches, search, expectation maximization, weighted backprojection, etc. Images may be stored in local memory and the reconstruction techniques applied by a local processor, or the images may be communicated to another device with separate memory and / or processor.

[0043] In some examples, one or more pillars of a multi-pillar sample may be imaged in a helical trajectory with combined beam deflection and translation stage rotation, as disclosed in the application "HELICAL TRAJECTORY VIA COMBINATION OF IMAGE SHIFT AND HARDWARE ALPHA TILT," filed March 23, 2021, the contents of which are incorporated herein by reference. Many of the disclosed steps may be automated by a controller, such as, for example, imaging a pillar, rotating to a subsequent angular position, re-imaging, translating to a new pillar, and imaging the new pillar.

[0044] Table 1 below shows the various sample processing time improvements that can be achieved using the multi-pillar example. Generally, throughput increases of 1.5x, 2x, 5x, 10x, 20x, or more can be achieved by using a multi-pillar sampler instead of a single pillar. [Table 1]

[0045] Preparing and imaging many pillars together reduces the time required for freezing the sample for tomographic image acquisition, transferring the sample between tools, and preparing the tool (pump-down, etc.). Thus, as shown, switching to 10 pillars per carrier improves processing time by a factor of 2.62.

[0046] FIG. 18 illustrates a multi-pillar sample 1800 including multiple pillars 1802a-1802e secured to a surface 1804 of a carrier 1806, for example, through a lift-out and welding process. In some examples, the carrier 1806 may instead be a separate substrate that can be attached to the carrier. The pillars 1802a-1802e may be arranged parallel to one another to facilitate tomographic imaging of each pillar. In some examples, the pillars 1802a-1802e are aligned in a line 1807 parallel to the surface 1804. In selected examples, the surface 1806 may be elevated at an angle relative to another reference plane of the carrier 1806, such as the bottom surface 1808, a reference plane of a separate carrier to which the carrier 1806 is attached, so that the line forms a small angle. In further examples, the pillars 1802a-1802e may be arranged on separate surfaces, such as in a stepped arrangement, to provide the line 1807 at a small angle.

[0047] 19 illustrates a multi-pillar sample 1900 including multiple pillars 1902a-1902c extending from respective substrates 1904a-1904c. The substrates 1904a-1904c are mounted on a carrier 1906, which may also, in some examples, be a substrate mounted on a separate carrier. The pillars 1902a-1902c may be arranged to form a line 1908 that is parallel to or at a small angle with respect to a reference surface 1910 of the carrier 1906. The small angle may be formed in a variety of ways, such as by a height difference between the different substrates 1904a-1904c, their relative placement on the different substrates 1904a-1904c, or by a step or angle in the reference surface 1910.

[0048] 20 is another method 2000 of forming a multi-pillar sample. In 2002, a plurality of pillars are formed from one or more substrate or sample materials, for example, using focused ion beam milling. In 2004, the pillars are attached to a substrate (which in some examples may correspond to a sample carrier) to form a multi-pillar sample, for example, using attachment techniques such as a micromanipulator tool and microwelding.

[0049] 21 is a further method 2100 of forming a multi-pillar sample. In 2102, one or more substrates are formed with one or more pillars attached to each substrate. Then, in 2104, the substrates with the attached pillars are positioned adjacent to one another on a substrate or sample carrier to form a multi-pillar sample.

[0050] Additional Examples The multi-pillar tomographic sample may include a bulk substrate and a plurality of spaced-apart tomographic pillars extending from the substrate and positioned at different positions along the substrate in a sequential manner. The method may include imaging a first tomographic pillar of the multi-pillar sample with an imaging beam, moving the multi-pillar sample using a motion stage such that a second pillar of the multi-pillar sample is in the field of view of the imaging beam, and imaging the second tomographic pillar with the imaging beam. The method may further include milling the bulk substrate to form multiple pillars. The apparatus includes an imaging system configured to direct the imaging beam at the multi-pillar sample and a motion stage configured to move and rotate the multi-pillar sample about a plurality of different pillar axes of the multi-pillar sample.

[0051] Further examples include those described in the following numbered paragraphs. 1. A multi-pillar tomographic sample comprising a bulk substrate and a plurality of tomographic pillars extending from the bulk substrate, spaced apart, and positioned at different locations along the substrate in succession. 2. The multi-pillar sample of the preceding paragraph, wherein the spaced apart successive ones form a line, and the line and spacing of the spaced apart successive ones are configured to reduce interference in tilt rotational imaging of each pillar. 3. The multi-pillar sample of any one of the preceding paragraphs, wherein the bulk substrate extends in a sample plane and the plurality of pillars extend in pillar planes parallel to the sample surface, parallel to each other and to the sample plane. 4. The multi-pillar sample of any one of the preceding paragraphs, wherein the bulk substrate extends in a sample plane and the pillars extend parallel to each other and to the sample plane, in pillar planes that have an oblique relationship to the sample plane. 5. The multi-pillar sample of paragraph 4, wherein the transmission cross sections of adjacent pillars are arranged in a non-overlapping spaced apart relationship in a direction perpendicular to the sample plane. 6. The multi-pillar sample of any one of paragraphs 2-5, wherein the spaced succession forms a plurality of spaced apart lines, forming a two-dimensional grid arrangement of pillars. 7. The multi-pillar sample of any one of the preceding paragraphs, wherein the pillars comprise cryogenic electron tomography pillars. 8. The multi-pillar sample of any one of the preceding paragraphs, wherein the pillars have a tip thickness less than about 400 nm, a base thickness greater than about 400 nm, and a length greater than about 1 μm. 9. A sample carrier comprising a carrier base and a multi-pillar sample of any of the preceding paragraphs supported by the carrier base. 10. A method comprising focused ion beam milling a raw bulk substrate to form the multi-pillar sample of any one of the preceding paragraphs. 11. A method comprising: imaging a first cross-sectional pillar of a multi-pillar sample with an imaging beam; moving the multi-pillar sample using a translation stage so that a second pillar of the multi-pillar sample is in the field of view of the imaging beam; and imaging the second cross-sectional pillar with the imaging beam. 12. The method of paragraph 11, wherein imaging the first tomographic pillar includes imaging at a plurality of rotational positions by rotating the multi-pillar sample about a translation stage rotation axis aligned with a first pillar axis of the first pillar, moving the multi-pillar sample includes moving the multi-pillar sample such that a translation stage rotation axis is aligned with a second pillar axis of the second pillar, and imaging the second tomographic pillar includes imaging the second pillar at a plurality of rotational positions by rotating the multi-pillar sample about a translation stage rotation axis aligned with the second pillar axis. 13. The method of paragraph 11 or 12, further comprising translating the multi-pillar sample along the first pillar axis to rotationally image another section of the first pillar before moving the multi-pillar sample to image the second pillar. 14. The method of any one of paragraphs 11-13, wherein imaging the first tomographic pillar includes rotating the multi-pillar sample to a +90 degree position or a -90 degree position to generate an unobstructed image at a position selected based on the angled linear arrangement of the multiple pillars of the multi-pillar sample. 15. The method of any one of paragraphs 12-15, wherein the multi-pillar sample comprises the sample of any one of paragraphs 1-10. 16. The method of paragraph 15, wherein the spaced-apart succession of multi-pillar samples forms a line, and the line and spacing of the spaced-apart succession are configured to reduce interference in tilt rotational imaging of each pillar. 17. An apparatus comprising: an imaging system configured to direct an imaging beam onto a multi-pillar sample; and a translation stage configured to translate and rotate the multi-pillar sample about a plurality of different pillar axes of the multi-pillar sample. 18. The apparatus of paragraph 17, further comprising a processor and memory coupled to the sample stage and imaging system, the memory including code that, when executed by the processor, causes the sample stage to rotate the multi-pillar sample about an axis of rotation aligned with a first pillar axis of the first pillar, move the multi-pillar sample such that the sample stage axis of rotation is aligned with a second pillar axis of the second pillar, rotate the multi-pillar sample about the second pillar axis, and cause the imaging system to direct an imaging beam through the first pillar at a plurality of rotational positions about the axis of rotation aligned with the first pillar axis and direct an imaging beam through the second pillar at a plurality of rotational positions about the axis of rotation aligned with the second pillar axis. 19. The apparatus of paragraph 17 or 18, further comprising a focused ion beam system configured to direct the focused ion beam to mill the raw sample substrate to form a multi-pillar sample. 20. The apparatus of any one of paragraphs 17 to 19, wherein the imaging system and translation stage include a transmission electron microscope.

[0052] General theory As used herein, image, projection, and image projection can refer to the distribution of electron or other imaging beam energy directed to and / or received by and detected by an imaging sensor after transmitting the imaging beam through a sample, presenting image data on a display, or presenting data associated with the field of view image stored in a JPG, TIFF, or other data file. While the disclosed examples are described with reference to electron beams as the penetrating energy used in 3D tomographic imaging, it will be understood that in some examples, other charged particle beam or energy beam sources, including x-rays and ultrasound, may be used. While the example microscopes are generally described with selected components for convenience in illustrating and explaining the principles of operation, it will be understood that other components may be included or removed.

[0053] The disclosed systems, methods, and apparatuses are not limited to any particular aspect or feature or combination thereof, nor do they require that any one or more particular advantages be present or problems be solved. While any theory of operation is provided for ease of description, the disclosed systems, methods, and apparatuses are not limited to such theory of operation. While some operations of the disclosed methods are described in a particular order for convenient presentation, it should be understood that this description style encompasses reordering unless a specific order is required by specific terms described below. For example, operations described sequentially may, in some cases, be reordered or performed simultaneously. In particular, many of the described image acquisitions and stage rotations and / or beam deflections need not be performed in a particular order. Some sequences, such as a series of images formed by a series of rotational increments or beam deflections at selected rotational increments, are generally preferred. Furthermore, for simplicity, the accompanying figures may not show the various ways in which the disclosed systems, methods, and apparatuses can be used with other systems, methods, and apparatuses. Additionally, this specification sometimes uses terms such as "generate" and "provide" to describe the disclosed methods. These terms are high-level descriptions of actual operations that are performed. The actual operations that correspond to these terms will vary depending on the particular implementation and will be readily discernible to one of ordinary skill in the art.

[0054] In view of the numerous possible embodiments to which the principles of the disclosed technology may be applied, it should be recognized that the illustrated embodiments are merely representative and should not be construed as limiting the scope of the present disclosure. The alternatives specifically addressed in these sections are merely exemplary and do not constitute all possible alternatives to the embodiments described herein. For example, various components of the systems described herein may be combined in function and use. We therefore claim the full scope of protection that falls within the scope of the appended claims.

Claims

1. providing a multi-pillar sample including at least a first pillar and a second pillar parallel to the first pillar; directing a charged particle beam at the first pillar; imaging the first pillar at a plurality of rotational positions by rotating the multi-pillar sample about a first pillar axis of the first pillar; directing the charged particle beam toward the second pillar; imaging the second pillar at a plurality of rotational positions by rotating the multi-pillar sample about a second pillar axis of the second pillar.

2. The method of claim 1 , further comprising reconstructing a 3D image of the first pillar and the second pillar.

3. The method of claim 2 , wherein the 3D image covers the entire angular span of the first pillar and the second pillar.

4. The method of claim 1 , wherein the range of the plurality of positions is at least 80 degrees.

5. 5. The method of claim 1, further comprising translating the multi-pillar sample along the first pillar axis to rotationally image another section of the first pillar before moving the multi-pillar sample to image the second pillar.

6. 2. The method of claim 1 , wherein imaging the first pillar comprises rotating the multi-pillar sample to a selected position of either a +90 degree position or a −90 degree position, and generating an image at the selected position having reduced obstruction by the second pillar or other pillars of the multi-pillar sample based on an angled linear arrangement of pillars of the multi-pillar sample relative to a 0 degree position.

7. 10. The method of claim 1, wherein the multi-pillar sample comprises a substrate defined by a length, a width, and a height, and wherein the first pillar and the second pillar extend from the substrate along the height of the substrate, are spaced apart, and are successively positioned at different locations along the length of the substrate.

8. The method of claim 7 , wherein the substrate is a sample carrier.

9. The method of claim 7 , further comprising attaching a plurality of pillars to the substrate to form the multi-pillar sample.

10. The method of claim 1 , wherein the distance between the first pillar and the second pillar is greater than 10 times the diameter of either the first pillar or the second pillar.

11. The method of claim 1 , further comprising directing a focused ion beam to mill a raw sample substrate to form the multi-pillar sample.

12. 1. An apparatus comprising: an imaging system configured to direct a charged particle beam to a multi-pillar sample including at least a first pillar and a second pillar; a translation stage configured to translate and rotate the multi-pillar sample about a plurality of different pillar axes of the multi-pillar sample; a processor and memory coupled to a sample stage and the imaging system, the memory comprising code that, when executed by the processor, causing the sample stage to rotate the multi-pillar sample about a first pillar axis of the first pillar and rotate the multi-pillar sample about a second pillar axis of the second pillar; a processor and memory that cause the imaging system to direct an imaging beam through the first pillar at a plurality of rotational positions about the first pillar axis and to direct the imaging beam through the second pillar at a plurality of rotational positions about the second pillar axis.

13. The multi-pillar sample is a substrate defined by a length, a width, and a height, the substrate extending in a plane defined by the length and width; 13. The apparatus of claim 12, wherein the first pillars and the second pillars extend parallel to one another from the substrate along the height of the substrate, are spaced apart, and are successively positioned at different locations along the length of the substrate.

14. 14. The apparatus of claim 13, further comprising a sample carrier, wherein the substrate is the sample carrier, and the first pillar axis and the second pillar axis form a pillar plane that is not parallel to a plane defined by the length and width of the substrate.

15. The device of any one of claims 12 to 14, wherein the first pillar and the second pillar have a tip thickness of less than 600 nm, a base thickness of more than 400 nm, and a length of more than 1 μm.

Images